A peer-reviewed study found serious adverse events exceeded COVID-19 hospitalizations prevented in the Pfizer and Moderna trials. Four years later, the standard rebuttal is still to cite a model that assumes vaccine harm is zero.
Yesterday, we learned how federal prosecutors in Maryland have criminally charged David Morens, a former senior scientific advisor to Anthony Fauci at the National Institutes of Health, with hiding records about COVID-19’s origins from the public. The indictment, unsealed this morning by the Department of Justice, alleges conspiracy, destruction of federal records, obstruction, and making false statements — all tied to communications Morens allegedly moved off official systems and deleted to evade FOIA requests.
He deleted them after a colleague warned him that FOIA was “just waiting to pounce.”
That story will keep. This one has been waiting four years.
Joseph Fraiman is an emergency physician in New Orleans. In 2022 he led a reanalysis of the Pfizer and Moderna phase III trial data — the same trials that authorized the mRNA COVID vaccines — and published the results in the peer-reviewed journal Vaccine. The finding was specific and sourced to the sponsors’ own submitted data: serious adverse events of special interest occurred more frequently in the vaccinated groups than in the placebo groups, at a rate that exceeded the reduction in COVID-19 hospitalizations within the trial window. In the Pfizer trial, the ratio was 4.4 serious adverse events for every hospitalization prevented. In Moderna’s, 2.4 to 1.
That paper has been in the public record since September 2022. It has not been retracted. It has not been refuted by a competing reanalysis of the same trial data — because Pfizer and Moderna have not released the individual participant data that would allow one. What it has received instead is the treatment Fraiman documents in meticulous and damning detail at Brownstone Institute this week: a BBC Radio 4 broadcast that brought in a reproductive immunologist to adjudicate a clinical trial reanalysis, let her make a false claim about what the paper says, and never checked whether the claim was true.
The false claim was this: that the authors had been “specifically told” the paper should not be used the way Dr. Aseem Malhotra was using it. No one told them any such thing. The peer review correspondence is publicly archived on Zenodo alongside the adjudication records and study data. The paper contains no such instruction. Anyone could have verified this in the time it takes to read eight open-access pages. The BBC reporter did not.
I want to add something to Fraiman’s account that his position as a physician-author prevents him from saying as directly as it deserves to be said, and that requires a different analytical lens than the one he’s using.
The methodological move that keeps appearing in mainstream coverage of this paper — and that appeared again in the BBC segment — is to counter a randomized controlled trial’s safety signal with a modeling study estimating that the vaccines saved millions of lives. Fraiman addresses this carefully. The logical problem deserves to be stated without softening.
The model they cited to rebut the trial cannot, by design, find the harm the trial found. That is not a rebuttal. It is a construction choice.
A randomized trial is the only study design that controls for the systematic differences between people who choose to get vaccinated and people who don’t. Those differences are large, consistent, and well-documented. People who proactively seek vaccination are, on average, healthier, more health-conscious, more likely to seek medical care when something goes wrong, and less likely to die from almost any cause in the near term — independent of the vaccine. This is the healthy vaccinee effect, and it inflates every observational efficacy estimate in the post-authorization literature. The modeling studies that estimate millions of lives saved are built on those observational inputs. They are not independent confirmation of the trial results. They are amplified versions of the same biased signal, processed through a mathematical structure that, as Fraiman notes, contains no term for vaccine-caused harm by design.
You cannot use a zero-harm model fed by healthy-user-inflated inputs to refute an excess harm signal in the sponsor’s own randomized data. The logical structure doesn’t hold. The randomized trial controls for the confounding the observational study cannot. When the two conflict, the trial wins — not because randomized trials are infallible, but because the specific failure mode of observational data runs in the direction of overestimating benefit and underestimating harm, which is exactly what’s at issue here.
Physicians and researchers who’ve written in after the earlier pieces on trial data suppression and the Cochrane review methodology failures already know what Fraiman is describing — they’ve watched the same pattern operate from inside institutions. The mechanism is always identical: when randomized data produces an inconvenient safety signal, the response is not to run a better randomized analysis. It is to cite a model that cannot, by construction, confirm the signal, and present the model’s failure to find it as exculpatory.
Readers who followed the analysis of VAERS signal suppression published here will recognize the second layer as well. After Fraiman’s paper was published, the journal Vaccine published two commentaries criticizing it — without notifying the authors in advance, without inviting a response, and then rejecting the authors’ submitted response letter without peer review. A journal that publishes criticism of a paper it accepted and then refuses to publish the authors’ reply is not conducting scholarly exchange. It is managing a narrative. The same paper was labeled misinformation on social platforms. No peer-reviewed study reporting favorable vaccine outcomes has received that label, regardless of methodology. The data-laundering pattern I documented in the Cochrane review analysis runs through the same infrastructure — different paper, identical institutional response.
There is a third layer the BBC broadcast missed entirely. Fraiman’s co-authors called on Pfizer and Moderna in an open letter in The BMJ, on the day of publication, to release the individual participant data. They have not done so in four years. The critics who insist that absence of participant-level data is fatal to Fraiman’s reanalysis have been remarkably untroubled that the same data remain withheld by the sponsors themselves. If the argument is that no one should draw conclusions from the public serious adverse event tables because the full data would be more informative, the implication is that no one — including regulators and the public — should be confident in the current harm-benefit picture until those data are released. That is not a position the critics appear willing to hold.
The BBC is not a fringe outlet. A failure of basic journalistic practice there is not a fringe problem. The host of a podcast called Everything Is Fake and Nobody Cares told his audience that Dr. Malhotra was spreading false information, on the basis of a claim about a paper he had not read, adjudicated by an expert who lacked the relevant methodology training, without disclosing that his chosen expert’s salary support flows from organizations founded by or formally partnered with the pharmaceutical industry. He produced, inside a broadcast about why fakery is no longer punished, a specimen of exactly that phenomenon.
Fraiman closes by noting that the host is half right: everything is fake, and some people do care. The individual participant data from the Pfizer and Moderna trials would settle the empirical question within months. That data has not been released in four years. Everything else in this story follows from that single fact, and it is the one fact that did not appear in the BBC broadcast.
What I haven’t laid out yet is how the journal’s coordinated refusal to publish the authors’ response connects to a documented pattern across at least four other vaccine safety papers where the same sequence occurred — submission, rejection without review, silence. That is the next piece. The pattern is clear enough now that the null hypothesis is becoming difficult to defend.
Fraiman’s full account: Brownstone Institute, April 27, 2026.
IPAK-EDU is grateful to Popular Rationalism as this piece was originally published there and is included in this news feed with mutual agreement. Read More
Leave a Reply